A Technical Review of EUS-Guided Variceal Eradication

ABSTRACT

Gastric varices present a unique therapeutic challenge for endoscopists. Although the use of direct endoscopic cyanoacrylate glue injection is superior to band ligation, it carries a risk of systemic adverse events. This led to the development of endoscopic ultrasound (EUS)-guided therapy. EUS enables accurate measurement and targeting of vessels, allowing for the precise intravascular delivery of cyanoacrylate and/or coils. Doppler imaging can be used to confirm obliteration in real time. In this review, we highlight recent literature on varying embolization techniques and detail the technical considerations required for a successful EUS-guided approach.

INTRODUCTION

Clinically significant portal hypertension can lead to the development of portosystemic collaterals in the gastrointestinal lumen (1). The location, supply, and drainage of these varices dictate management options. The Sarin classification further characterizes the endoscopic location of gastric varices and helps guide therapy (2). These locations have important clinical implications when deciding management approaches. Esophageal varices are more common and frequently observed, whereas gastric varices tend to be more severe, with a higher risk of recurrent bleeding and increased mortality rates (3–5).

Hence, understanding the endoscopic management options for gastric varices is important. In this updated review, we will outline the relevant literature that culminated in the use of combination endoscopic ultrasound (EUS)-guided therapy for gastric variceal obliteration. It is important to highlight that these therapies should be used for bleeding gastric varices; data regarding primary prophylaxis are limited. Next, we will focus on the technical aspects (including our own anecdotal practices in this patient population).

ENDOSCOPIC GLUE INJECTION

Endoscopic intravariceal injection of glue has been widely practiced for the treatment for gastric varices since first reported in 1986 (6). Cyanoacrylate is a monomer that rapidly polymerizes when it comes in contact with an ionic medium such as blood, forming a hard substance. Numerous studies have validated glue as a reliable treatment option with a 93% hemostasis rate at 1 week (7). A single randomized trial (yet to be replicated) suggested cyanoacrylate superiority over beta-blockade for primary prophylaxis; however, in the absence of a standardized risk stratification system, it remains unclear which patients should undergo prophylaxis (8). The glue adheres to the vessel and rapidly precipitates blood clotting after injected. Extra precautions need to be taken to avoid contact of glue with the endoscope and its working channel. Unexpected adverse events have been linked to cyanoacrylate injection, including systemic embolism, cerebral stroke, needle impaction, variceal nest ulceration/bleeding, and even death (9–11). It is important to note that there is not a universal method for glue injection, and this may significantly affect the incidence and type of specific adverse events. For example, using glues with quicker polymerization formulas may increase risk of needle impaction, although additional studies are needed to confirm this association.

EUS-GUIDED CYANOACRYLATE GLUE INJECTION

In this context, the evolution and adaptation of EUS-guided endovascular therapies rapidly emerged. This paradigm shift was driven by safety concerns because less glue is needed to achieve obliteration when the feeder vessel is targeted. EUS also provides direct visualization of the vessel, which allows precise intravascular targeting and confirmation of varix obliteration in real time. EUS-guided glue injection was first reported in 2007 (12). Numerous comparative studies between EUS-guided and endoscopic glue injection have suggested that an EUS approach may reduce rebleeding rates (13–16). While effective, there are still adverse events (specifically embolic phenomenon) associated with EUS-guided cyanoacrylate glue. This highlights the importance of obtaining cross sectional imaging, where one can delineate the vascular anatomy. For instance, a patient with no portosystemic shunt cannot suffer from systemic embolization with endoscopic therapy negating this major risk. Furthermore, there is significant study heterogeneity and selection bias that must be considered when evaluating these outcomes (13–16).

EUS-GUIDED EMBOLIZATION COILS

The use of embolization coils under EUS guidance has offered endoscopists an additional tool with a different mechanism of action. The coils are stainless steel, synthetic fiber that serves as a hemostatic scaffold that obstructs blood flow (Figure 1). The size and number of coils used is based on the characteristics of the gastric varices (typically the size of the coil is 1.5 times the varix diameter) (17). Although the risk of coil migration is minimized through this approach, it has still been reported (18). Coiling was first reported in 2007 in a case report that described its use in treating ectopic varices (19). Its application for gastric varices targeting the feeder vessel was later described in 2010 (20). Since then, a multitude of studies have described the use of coiling as monotherapy or combination therapy with cyanoacrylate and other injectates. A recent multicenter study involving 106 patients examined technical and clinical outcomes at high-volume and low-volume centers across the United States (21). The study found that there was a high technical (100%) and clinical (89%) success rate with no differences based on the center's volume. They did note 2 intraprocedural (pulmonary embolism and variceal bleeding from needle access) and 5 post procedural adverse events (21).

Figure 1.

Nester embolization coils, 2 cm long with a 2 mm diameter.

EUS-GUIDED COMBINATION THERAPY

The therapeutic EUS-guided approach, combining coil and cyanoacrylate glue therapy, has been thoroughly studied and widely compared in the literature (22–27). Yet, such comparisons are limited by study heterogeneity such as injection methods and patient selection. A retrospective study comparing EUS-guided coiling (n = 11) vs cyanoacrylate injection (n = 19) found that although obliteration rates were similar (91% vs 95%), the coil group required fewer endoscopies and experienced fewer adverse events (9% vs 40%) (13). A large study of 152 patients with more than 6 years of follow-up demonstrated that EUS-guided coiling plus cyanoacrylate injection was technically effective (>99%) for active bleeding, primary and secondary prophylaxis (23). There were only 3 episodes of posttreatment bleeding in long-term follow-up. A 2020 randomized controlled study involving 60 participants further supported the use of combination therapy (28). In this trial, both approaches demonstrated excellent technical (100%) success rates; yet coiling plus cyanoacrylate injection proved superior in lower rebleeding (20% vs 3.3%) and reintervention rates. A recent randomized control study in 2025 also compared combination with monotherapy therapy in a cohort of 50 patients (26). The combination of coiling plus cyanoacrylate injection was associated with higher obliteration rates (100% vs 92%), fewer reinterventions (21% vs 54%), longer time to reintervention, and improvement in survival outcomes (26). A meta-analysis of 11 studies (536 patients) confirmed that EUS-guided combination therapy should be the preferred strategy over monotherapy (29). The authors found that adverse events were lower (10% vs 21%) in the combination groups. The management of gastric varices is very complex, and further studies with well-defined patient populations and treatment end points are needed to help define the clinical impact of combination therapy for secondary prophylaxis.

ADDITIONAL MODALITIES FOR TREATMENT

Absorbable hemostatic gelatin sponge

The use of an absorbable gelatin sponge (AGS) used in interventional radiology and surgical procedures has garnered attention in recent studies in combination with coiling (30–32). AGS is derived from purified porcine gelatin where it acts as a water-insoluble plug that can absorb 45 times its weight in blood (33). There is also a benefit of no damage to the endoscope or posttreatment ulceration. Studies have looked into this as an alternative to cyanoacrylate glue. A case series in 2020 looked at outcomes of EUS-guided coiling and AGS in 10 patients with either active bleeding or those requiring secondary prophylaxis (30). There were no instances of rebleeding or reintervention over a mean follow-up of 6 months, and on follow-up of EUS, there was near obliteration of varices. That same group later conducted a retrospective matched cohort study that showed combining EUS-guided coiling with AGS was superior to traditional endoscopic cyanoacrylate injection, resulting in lower rebleeding rates over a 9-month follow-up period (31). It is noteworthy that before injecting gel foam, coils were deployed to significantly reduce Doppler flow on EUS until an absence of contrast run-off on fluoroscopy was achieved, which required a mean of 8 coils per case. Hence, varix obliteration may have been achieved with coils alone, and the added benefit of adding AGS remains to be proven. AGS slurry is not currently FDA-approved, potentially restricting its use for now.

Thrombin injection

Thrombin injection has also been investigated as an alternative to glue. The use of thrombin is promising, owing to its excellent safety profile (less risk of systemic embolization and ulceration). Its mechanism of action relies on human-derived thrombin that converts fibrinogen to fibrin thereby promoting clot production (34). EUS-guided injection of thrombin was first reported in 2018 in a case series involving 8 patients (3 active bleeding and 5 primary prophylaxis); only 1 patient with active bleeding failed to achieve hemostasis (35). A feasibility, safety, and outcomes study evaluating EUS-guided thrombin injection and coil implantation for gastric varices in 20 patients undergoing 33 procedures showed a 95% technical success rate, 85% obliteration rate, and 2 recurrent bleeds on follow-up. A mean of 4 coils were implanted before thrombin injection. Two small, but randomized controlled trials comparing endoscopic thrombin with glue injection have shown similar hemostasis and obliteration rates, but lower rates of adverse events in the thrombin group, in particular no incidences of gastric ulceration (0% vs 37%) (36) and lower rates of rebleeding (37). Further randomized studies with larger sample size and longer follow-up are needed. Cost may be a limitation because human thrombin is much more expensive than glue.

TECHNICAL ASPECTS

Patient selection and preprocedural considerations

First and foremost, we always have a detailed multidisciplinary discussion with our transplant hepatology and surgical colleagues when selecting patients for EUS-guided variceal obturation. Understanding a patient's clinical prognosis and eligibility for alternative therapies such as transjugular intrahepatic portosystemic shunts placement and retrograde transvenous obliteration helps guide clinical decision-making. Preprocedure cross-sectional imaging (CT/MRI) is important to define the vascular anatomy (i.e., portosystemic shunt) and characterize the embolic risk. We then have a detailed discussion with the patient regarding the procedure risks and adverse events (detailed above). This procedure can be performed with monitored anesthesia care, and most studies to date used monitored anesthesia care. General anesthesia should be considered if there is a need to instill large volume of saline into the stomach to outline varices and differentiate intramural varices from extramural collaterals. In general, patients should receive preprocedural antibiotics (e.g., third-generation cephalosporins).

Procedural steps

A standard upper endoscope is initially used to characterize the size and location of the varices. We then proceed with a curvilinear echoendoscope (GF-UCT180; Olympus America Inc) and advance to the gastroesophageal junction or gastric cardia, and sometimes it is necessary to turn clockwise toward the fundus (spleen) to locate the varices (Video 1). Based on EUS imaging, we instill water into the lumen to help delineate the gastric wall and optimize visualization of the gastric varices (Figure 2). One must be careful not to excessively use water if the patient is not intubated to minimize risk of aspiration. Using Doppler, we carefully inspect for perigastric/intraabdominal varices (as they can sometimes be misinterpreted for gastric varices) and other vascular abnormalities (i.e., pseudoaneurysm). Once the gastric varices have been identified, we sometimes look for feeder vessels. Although targeting the feeder vessel can be more technical demanding, one study suggested it requires fewer coils and glue with comparable obliteration rates (38). A transesophageal or transgastric approach can be used (Figure 3). There are no data to support one method over the other.

Figure 2.

Endoscopic ultrasound view after water is instilled into the lumen to help delineate the gastric wall from varices.

Figure 3.

Endoscopic ultrasound-guided approaches to varix treatment include (A) transgastric puncture directly into the varix, (B) transesophageal access directly into the varix, and (C) transesophageal puncture targeting the feeder vessel.

Once a target has been identified, a 19- or 22-gauge fine needle aspiration (FNA) needle is chosen. We prefer the 19-gauge (e.g., Expect needle; Boston Scientific, Natick, MA) owing to the large 0.035” diameter coils that can be used. The stylet is removed, and the FNA needle is primed with normal saline if 2-octyl cyanoacrylate is used or D5W to prevent premature polymerization within the needle if N-butyl-2 cyanoacrylate is used. The needle is advanced into the targeted varix and either a small volume of saline or D5W injected or blood aspirated to confirm varix entry. If blood is aspirated, the needle should be flushed afterward with saline or D5W. The needle tip is advanced close to the opposing wall of the varix (care is taken to avoid placement of the coil into the opposite wall to allowing for coiling and minimizing risks of perforation and coil extrusion). The needle is then locked to secure it in place. The embolization coil (eg, Nester; Cook Medical, Bloomington, IN, USA) is then loaded with the stylet serving a pusher to the advance the coils. The coil is then deployed using the stylet to advance it in; on deployment, the coils will appear as bright hyperechoic rings on EUS (Figure 4). The FNA needle is flushed with saline or D5W to prevent clot formation. Based on real-time assessment of vessel obliteration, additional coils can be inserted in the same target without removing the needle.

Figure 4.

Endoscopic (A) and endoscopic ultrasound view of gastric varices (B). The varix is punctured with a needle (C); after combination therapy, real-time obliteration is seen with Doppler imaging (D).

If combination therapy is pursued, the next step would be cyanoacrylate glue injection through the same needle and puncture site. The volume of cyanoacrylate injection is based on the diameter of the targeted vessel. Larger (>2 cm) varices may require several ml of glue to achieve obliteration. Using 2-octyl cyanoacrylate, glue is injected slowly through the catheter over 45 seconds. D5W saline is used to flush any residual glue in the needle into the varix. The needle is then retracted into the sheath, the sheath advanced a distance beyond the tip of the echoendoscope, and the echoendoscope with fully inserted sheath removed from the patient. The sheath tip can then be cleaned of any residual glue before removing the needle through the working channel. A similar but slightly alternative technique is detailed in Supplemental File 1 with video (see the Supplementary Digital Content, http://links.lww.com/CTG/B466).

The use of fluoroscopy during treatment is optional. Contrast can be injected before treatment to define feeder vessels, flow direction, and presence of shunts. Glue can be mixed with the appropriate contrast agent such as ethiodized oil (Lipiodol), but this will increase the viscosity of the injectate. We do not use fluoroscopy because we have not found it to significantly alter our endoscopic approach.

As detailed earlier in this article, EUS-guided combination therapy with coiling and glue is commonly used. However, due to the side effect profile of cyanoacrylate, alternative methods have been studied. At the University of North Carolina, we started to use a novel hemostatic gel (PuraStat; 3-D Matrix, Newton, Massachusetts, USA) as a substitute for cyanoacrylate (39). The coils serve as a scaffold for the gel to concentrate and exert its effect in the vessel. We inject 3 mL of Purastat through the needle immediately after placing the last coil. Its ease of use and safe profile simplifies the procedural steps for our staff and endoscopy time. These data are preliminary and present an area for future prospective study/case series.

Postprocedural care

After the varix is treated, we often repeat surveillance endoscopy with EUS every 3–6 months until eradication is achieved. One can also consider repeat endoscopy and/or cross sectional imaging 4–6 weeks after treatment (especially in cases where active bleeding was treated). We tailor our therapies each session based on the size of varix and location of previously placed coils.

Procedure-related costs

The costs associated with EUS-guided gastric varix embolization is difficult to capture. There is heterogeneity across different hospital endoscopy centers, inpatient versus outpatient status, operator preferences (especially related to equipment), and reimbursement of alternative agents. Furthermore, hospitalizations for variceal bleeding are one of the more expensive gastrointestinal-related disease processes (40). The associated rebleeding and mortality rates are as high as 40% and 50%, respectively (41). An accurate cost-effectiveness study to compare EUS therapies versus direct endoscopic injection versus endovascular therapy is needed. Such a study can also help describe secondary costs related to readmissions, rebleeding, and repeat procedures.

When using this technique at your center, the Current Procedural Terminology code used for EUS-gastric varix injection therapy is ultrasound-guided transmural injection of diagnostic or therapeutic substance(s), 43253.

CONCLUSION

The management of gastric varices has evolved significantly with the emergence of EUS-guided therapies. Among available options, combination treatment with embolization coils and cyanoacrylate glue demonstrates the most promising outcomes in safety, efficacy, and long-term durability. Novel agents such as AGS and thrombin show potential, particularly in minimizing adverse events, although further validation studies are needed to determine the added benefit to coil implantation alone. A tailored, multidisciplinary approach is essential in optimizing care for patients with gastric varices. As techniques advance and more data become available, EUS-guided interventions are likely to become the cornerstone of variceal management.

CONFLICTS OF INTEREST

Guarantor of the article: Andrew Canakis, DO.

Specific author contributions: A.C.: wrote the draft of the manuscript. K.B., and T.B.: edited and approved the final manuscript.

Financial support: None to report.

Potential competing interests: Dr. Baron is a consultant for Boston Scientific, ConMed, W.L. Gore, Cook Endoscopy, and Olympus America. Dr. Canakis and Binmoeller declare no relevant funding for this work or financial relationships.

Supplementary Material

ct9-17-e00983-s001.docx ^{(16.2KB, docx)}

ABBREVIATIONS:

AGS

absorbabale gelatin sponge

EUS

Endoscopic ultrasound